Report from RSNA: Industry ties to research pose ethical questions

December 9, 2004

One-fifth of the abstracts presented at the 2003 RSNA meeting had at least one author who disclosed ties to industry. Such papers were twice as likely as those without any disclosures to discuss an unlabeled use of a commercial product, according to a presentation from Harvard University Medical Center.

One-fifth of the abstracts presented at the 2003 RSNA meeting had at least one author who disclosed ties to industry. Such papers were twice as likely as those without any disclosures to discuss an unlabeled use of a commercial product, according to a presentation from Harvard Medical School.

"This is not surprising, nor is it necessarily inappropriate for commercial interests to influence the topics of radiological research. However, studying unapproved uses of a product may not be the best use of research money and may not answer the most important questions about the use of a product," said lead author Dr. Stephen Brown.

Furthermore, studying non-FDA-approved uses may serve to avoid unfavorable issues such as complications associated with the approved use of a product. It may also avoid comparing the product's approved uses against other products that may be superior, Brown said.

He noted that close ties between industry and academia may promote research designed to maximize commercial benefits rather than research to determine fundamental importance and further long-term public interest.

Of 1549 abstracts published in the 2003 RSNA program book, 271 disclosed at least one author with a financial tie to a company whose products or services were covered in the report. Of these 271 abstracts, 97 had one author disclosure, 77 reported two, and 97 had three or more.

Of the 15 types of corporate disclosures reported, the most common were authors who were employees, those who received corporate grants, and those who were corporate consultants or shareholders.

More than 30% of papers disclosing a corporate relationship discussed the non-FDA-approved use of a commercial product, compared with 15% of papers with no disclosed ties to industry. The difference was significant.

These data do not prove that bias occurred in the selection of research topics. Their exposure, however, raises questions that merit further attention:

  • How often and to what degree were financial interests given priority over those of academia and the public when decisions were made regarding the selection of radiological topics by researchers?

  • Are researchers silently granted financial inducement to advertise non-FDA-approved uses of industry products?

  • Are companies who support the presentation of data on non-FDA-approved products doing so for marketing purposes?

  • Who has greater weight for establishing the research agenda for radiology: industry or academia?

Brown cited the need within radiology to scrutinize more closely the decision-making process to determine which products are studied and the properties for which they are evaluated.

He also advocated creating a greater dialogue between industry and academia to minimize the possibility that financial self-interest will improperly influence what products are tested and how the tests are conducted.

"These results suggest a blurring in academic radiology of the distinction between fundamental intellectual inquiry and commercially targeted development," he said.